I'd suggest making multiple PRs, one depending on the other one possibly, to each be committed with the appropriate attribution.
On Fri, Jan 27, 2017 at 5:05 PM, Ryan Merriman <merrim...@gmail.com> wrote: > What would our process be if someone did contribute a commit to a pull > request? > > On Fri, Jan 27, 2017 at 4:02 PM, Casey Stella <ceste...@gmail.com> wrote: > > > Yes, we should definitely not destroy authorship information. To my > > knowledge that hasn't happened yet and we should ensure it does not > happen > > in the future. > > > > On Fri, Jan 27, 2017 at 5:00 PM, P. Taylor Goetz <ptgo...@gmail.com> > > wrote: > > > > > IMO, that’s okay as long as the commits in a pull request are from a > > > single author. But it is possible for pull requests to contain commits > > from > > > multiple authors. If you squash those commits, you are potentially > > > destroying authorship information, which I would advise against. > > > > > > -Taylor > > > > > > > On Jan 27, 2017, at 4:51 PM, Casey Stella <ceste...@gmail.com> > wrote: > > > > > > > > Just so we're clear, we do squash commits upon merge (we followed the > > > suit > > > > of Apache Mahout and use --squash as described at > > > > https://mahout.apache.org/developers/github.html# > > merging-a-pr-yours-or- > > > contributors), > > > > but we do not merge commits from multiple people into a single > commit. > > > I'm > > > > guessing that's kosher, but it's something we probably should > clarify. > > > > > > > > On Fri, Jan 27, 2017 at 4:46 PM, P. Taylor Goetz <ptgo...@gmail.com> > > > wrote: > > > > > > > >> While it certainly doesn’t hurt to have one, it’s not strictly > > required. > > > >> It *is* required for committers though. > > > >> > > > >> When you merge a pull request, the authorship information is > > maintained. > > > >> Just make sure you don’t squash other people’s commits. > > > >> > > > >> -Taylor > > > >> > > > >>> On Jan 27, 2017, at 4:36 PM, Casey Stella <ceste...@gmail.com> > > wrote: > > > >>> > > > >>> Hi Mentors, > > > >>> > > > >>> I was wondering if you could help me settle a question. What is > the > > > >> ASF's > > > >>> stance on ICLAs for non-committer contributions? Are they > required? > > > >>> > > > >>> On the one hand, https://www.apache.org/dev/ > > committers.html#applying- > > > >> patches > > > >>> requires only that we attribute appropriately to form a legal > > > papertrail > > > >>> via the git history. Also, this discussion ( > > > >>> http://marc.info/?l=incubator-general&m=142175320215392&w=2) seems > > to > > > >>> indicate that they are not required. > > > >>> > > > >>> On the other hand, http://www.apache.org/licenses/#clas indicates > > that > > > >> the > > > >>> ASF "desires" ICLAs for contribution. I also see some projects > > > requiring > > > >>> them (i.e. flink and brooklyn) of contributors. > > > >>> > > > >>> Thanks in advance for the clarification! > > > >>> > > > >>> Best, > > > >>> > > > >>> Casey > > > >> > > > >> > > > > > > > > >
